COMPARABLE WORTH: SOME ISSUES FOR CONSIDERATION
SHARON BERNSTEIN MEGDAL*
The comparable worth approach is a radical departure from market determination of wages. This paper considers some eco- nomic issues pertinent to the comparable worth dialogue and focuses on implications of comparable worth legislation at the state and local levels. We show that current sex diflerentials in wages could rqlect voluntay hbor supply choices and/or dis- crimination. We also discuss the reverberations of comparable worth and note that some women in an occupation targeted with a wage increase could be made worse off as a re-sult of employer cutbacks. We discuss why state OT local officials may be hesitant to adopt comparable worth and also ofle~ possible explanations for their being receptive to it. The intent throughout is to provide a balanced d i s d n that will aid individuals in taking an edu- cated stand on the issue.
The comparable worth doctrine-that is, equal pay for jobs of compara- ble value-has received much attention in the popular and professional media. Under comparable worth payment schemes, wages are set accord- ing to a jobs requisite skills, effort, working conditions, and responsibility. Such schemes have been proposed and in some cases adopted to correct for perceived inequities in the market for labor services. The comparable worth approach is a radical departure from market determination of wages. Proponents of comparable worth argue that discrimination has pre- vented labor markets from assigning proper values to the output contribu- tions of women working in traditionally female occupations. Proponents give little credence to the arguments of comparable worth opponents, who contend that the invisible hand of a market economy should not and cannot be replaced by the visible hand of a wage-setting committee. Opponents of comparable worth make two basic points: (1) it is impossible for any wage evaluation committee to characterize the worth of jobs fully and consis- tently, and (2) more importantly, any compensation scheme which ignores market forces will cause tremendous economic distortions. Proponents, however, claim that policies and discriminatory actions of economic agents
Member, Arizona Corporation Commission (on leave from the Department of Eco- nomics, University of Arizona, kcson); member, America Economic Association Committee on the Status of Women in the Economic Profession. I would like to thank Paula Cech, Mari- anne Ferber, and Walter Oi for helpful comments on an earlier version of this paper.
Contemporary Policy Issues Vol. IV, April 1986
MEGDAL: COMPARABLE WORTH ISSUES 41
have created an economic system far removed from the ideal market econ- omy.' Because polemics often stand in the way of rational discourse, each side is skeptical of the other.
Although comparable worth is out of favor at the federal level and is rejected by most of the private sector, it receives much attention at the state and local levels. A number of states have implemented comparable worth pay rules, and other states are contemplating adopting such rules. Substate governments, such as cities and counties, also are studying the issue. This paper considers some economic issues pertinent to the comparable worth dialogue, and focuses on the implications of comparable worth legislation at the state and local levels.
In section 11, we show how predominantly female, low-wage occupa- tions can be observed in the absence of discrimination. Then we consider discrimination and the difficulty of distinguishing situations in which dis- crimination is present from those in which it is absent. (Presumably, com- parable worth advocates seek to rectify the former rather than the latter.) In section 111, we discuss the implications of comparable worth by tracing the likely responses of employers. Public-sector adoption of comparable worth legislation is discussed in section IV. We summarize and conclude in section V.
11. WAGE DETERMINATION AND DISCRIMINATION
Let us consider a simple economy in which two occupations exist. For purposes of our discussion, we will call them the predominantly female occupation (PFO) and the predominantly male occupation (PMO). Let US further assume that the two occupations require different skills. Therefore, one type of labor cannot be easily substituted for the other, and each labor market's equilibrium is at the intersection of a downward sloping demand curve and upward sloping supply curve.2
It is quite easy to describe a scenario in which the equilibrium wage of the PFO, w,, is below that of the PMO, w,. Let the first occupation follow a work schedule like that of public school systems, require considerable fin- ger dexterity, and allow for easy exit and entry. Let the other occupation involve a rigid 8 a.m. to 5 p.m. schedule 50 weeks per year, require physi- cal strength, and involve training of the same duration but of a different nature than the first occupation. Hence, without additional (unpaid) train- ing, individuals cannot switch occupations. In a society where women tend to assume a major share of child rearing and where day-care programs are limited, women likely would gravitate to the first occupation. Men would gravitate to the other occupation, if for no other reason than the first may
1. The literature on comparable worth is extensive. For arguments by economists, see
2 . We assume competitive labor markets exist and, therefore, we abstract from unions and Bergmann (1985), Killingsworth (1985), and Waldauer (1984).
42 CONTEMPORARY POLICY ISSUES
be viewed as womens work. If an employer has no preference as to male versus female employees-i.e., all that an employer cares about is whether an employee can do the job-natural sorting of workers into a PFO and PMO would result. Given this sorting, the ratio of wF to w, would be deter- mined by the occupations demand curves (which depend on labor produc- tivity and on prices of the output being produced), and by the occupations supply curves (which depend on the number of individuals supplying their services and the extent to which their willingness to work varies with the wage rate). It is possible for a ratio to be greater than, less than, or equal to one.
Thus, it is assumed that supply behavior depends on voluntary choices and that demand behavior is independent of the workers sex-that is, the employer cares only about a workers contribution to the value of output. Still, we could observe an equilibrium wage in the PFO well below that in the PMO, because the equilibrium wage depends on the relative positions of the labor demand and supply curves. These, in turn, depend on more than the skill, effort, and responsibility associated with a job.
If we define sex discrimination in employment as differential wage- setting, hiring, or promotion practices based solely on the sex of the worker, the situation described above does not appear discriminatory. The PFO carries a low value, or low wage, because of labor-supply choices and employer demand. Some who deny that labor-market discrimination exists argue that current wages reflect voluntary decisions; women tend to choose certain occupations over others. In addition, such critics note that not only is it illegal for an employer to discriminate, but that doing so does not make economic sense. If a woman is willing to work at a lower wage than is an equally productive man, the employer will hire the woman. Profits will be maximized for a given output level when costs are minimized. Thus, there may be nothing evil regarding an employer who hires women to perform a task-thereby making it a predominantly female job-if women are will- ing to work at lower wages than are equally productive men. Likewise, it does not make economic sense for an employer to pass over a qualified female and promote or hire a less-qualified male.3
So, if it is both illegal and economically foolish to discriminate, how can we argue that the wage structure reflects discrimination? If women are willing to work at lower wages than are equally productive men, shouldnt an employer take advantage of the situation? Those who defend the current wage structure would argue that in fact there is no problem because (1) discrimination is illegal; (2) women, for various reasons, are willing to work for lower wages; and (3) employers are profit maximizers. Yet many,
3. We should acknowledge situations in which discrimination could be considered rational. Some argue that it might be rational for an employer to discriminate if male workers require a sa lary premium to work alongside female workers. Others cite statistical discrimina- tion as causing sex bias in hiring and promotion.
MEGDAL: COMPARABLE WORTH ISSUES 43
including comparable worth opponents, admit that the existing wage structure reflects discrimination. Can these two positions be reconciled?
They can be if the definition of discrimination is broadened to include employer behavior which is based on incorrect perceptions of productivity differences by sex, and worker behavior which is based on past and present barriers to entry. The incorrect perceptions could result from failing to observe female productivity and/or from unjust presumptions. An exam- ple of the latter is when an employer is unwilling to hire a female manager because it is believed that male workers will not recognize her authority. Another example is when an employer is unwilling to hire a particular woman because other women have left the company to raise families. (The latter is an example of what is called statistical discrimination.) Barriers to entry include exclusionary practices of employers, schools, and training programs. Sex-stereotyping behavior of parents, teachers, guidance coun- selors, and the media also can dissuade women and girls from entering cer- tain occupations.
Another definition of sex discrimination in employment, then, is any sex- based difference in employment outcome that would not occur in a gender- neutral society. Until sex discrimination in this broader sense is eliminated, occupational choices and the existing wage structure will not reflect free choice an the part of all workers.
We can examine the implications of discrimination graphically by using a simple model of wage determination. First, consider the situation in which an employer is hiring for a position and perceives no differences in productivity by sex. In panel (a) of figure 1, we show the supply curve of women to this occupation (SF), and the supply curve of men to this occupa- tion (SM). To obtain the supply curve representing the total number of peo- ple willing to work at each wage, we combine the number of women and men willing to work at that wage. Panel (b) shows the resulting total supply curve (ST). The equilibrium wage (w*) is determined by the intersection of the employers demand curve (D) and S,. At w*, L* workers are hired, L,* of whom are female. The remaining employees (LG = L* - L$ are male. All workers are paid the same wage.
Now, consider a situation in which women will not be hired to fill a cer- tain position because the employer incorrectly perceives differences in pro- ductivity per unit of labor. Essentially, laborers are divided into two p00ls.4 The employers incorrect perceptions effectively serve as a barrier to entry. When women cannot enter certain occupations, they crowd
4. This situation once existed in some segments of the insurance industry, where women and men were segregated into in-office and out-of-office sales personnel, respectively. Women were effectively barred from holding a job for which they were qualified. Although such bla- tant examples of this type of discrimination are becoming less prevalent, such employer behav- ior clearly imposes a barrier to womens entry into the affected occupation.
44 CONTEMPORARY POLICY ISSUES
into occupations which are open to them (see Bergmann 1971). This crowd- ing results in lower wages in occupations open to women, and higher wages in occupations from which women are excluded. Figure 2 depicts the mar- kets for two occupations. Assume that in the absence of discrimination, the demand and supply curves are D,, Dp, Sly and S2, and that the equilibrium wage is the same in both occupations (w: equals w:). Each supply curve can be thought of as the sum of the supply of men and women to the occu- pation at each wage.
Now, let us introduce discrimination. Assume that women face discrimi- nation in that they are crowded out of occupation 1 and crowded into occu- pation 2, where we implicitly assume few men are working. As a result, the female supply curve to occupation 2 shifts to the right, and the female sup- ply curve to occupation 1 is eliminated. Thus, the immediate impact of the crowding is to shift occupation 1s overall supply curve to the left and to shift occupation 2s supply curve to the right. Let S ; and S; represent the resulting supply curves. The shifts put upward pressure on the wage in occupation 1 (the wage rises to w:), and put downward pressure on the wage in occupation 2 (the wage falls to w;). Because supply is dependent on the wages in alternative occupations, changes in wages will induce changes in male labor supply behavior. (Although women have no alterna- tive employment in this example, the aggregate female supply curve could change as their husbands earnings change.) These second-order changes will not likely be enough to offset the original effect of discrimination, which is to increase the wage in occupation 1 relative to that in occupation 2.
MEGDAL: COMPARABLE WORTH ISSUES
Many argue that when a certain occupation becomes predominantly female, it consequently becomes a low-wage occupation. That is, because women perform the job, it is less valued by society (bank telling is often used as an example), This discussion demonstrates that the decrease in wages occurring when an occupation becomes predominantly female may be a direct result of barriers to entry existing elsewhere in the economy, rather than a reflection of low value assigned to womens work. Employ- ers could not hire women at a low wage unless women were willing to work at that wage, but their willingness to work at a low wage may result from the crowding phenomenon.
When discrimination crowds women into a few occupations, the observed wage structure does not reflect voluntary choices. Therefore, it is not the wage structure associated with an unhindered free market econ- omy. Furthermore, because the prevailing wage structure and pattern of occupational segregation depend on supply and demand, the extent to which these schedules reflect discrimination determines the extent to which current wages reflect discrimination.
Most observers agree that the relative position of female labor has improved as discriminatory practices have been curtailed. The c...